For the estimated 30 million uninsured Americans, navigating insurance enrollment can be confusing, cumbersome, and intimidating. The hassles are so great that some people have avoided enrollment altogether, and that can be costly for everyone. And when the Biden administration lifts the COVID-19 public health emergency, as it plans to in April, another 18 million people could lose Medicaid coverage.
But HKS Associate Professor Mark Shepard says there is good news for the insurance enrollment process. And he has the data to back it up. We spoke with Shepard about his new working paper, “Reducing Ordeals Through Automatic Enrollment: Evidence from a Health Insurance Exchange.”
Q: How did this research come about?
I’ve been working for a number of years on research, studying the Romneycare Massachusetts Health Insurance Exchange, because, believe it or not, it’s still the best source of data for a health insurance exchange for low-income Americans.
I saw this interesting policy change that happened: The state auto-enrolled the poorest individuals in the Massachusetts Exchange who qualified for free coverage to reduce enrollment frictions. Then in 2010, they had to cut auto-enrollment for financial reasons, requiring users to choose a health plan. And it just jumped out immediately, in just the raw statistics, what a big difference this makes on the number of people enrolling in insurance plans. That extra step of choosing a reduced enrollment plan in the program by one third, which was a huge impact. It was the equivalent impact that we’ve seen, from other work, of the effect of raising premiums by almost 60%.
The big question guiding my research is, how do we make health insurance markets work? We know a lot from economics about how markets and competition can work well to benefit society. But we also know a lot from economics about how insurance markets—and health insurance markets in particular—fail in big ways.
So, I think the goal of this line of research is to better understand what policy strategies can be put in place to have not just laissez-fair market competition, but regulated market competition that would yield some of the benefits of choice and competition that we know about from economics, while safeguarding against some of the flaws and problems that we know arise in insurance markets.
Q: What did your study reveal about insurance enrollment?
You could imagine designing some social programs where everyone qualifies. There would be some universal programs, say public education. Everyone qualifies for public education for their kids.
But other programs are targeted, and they’re targeted usually at low income people or certain demographic groups who really need it. The problem with that is it puts the onus on the beneficiary to prove that they qualify.
There were two main findings in our study. One was that even a pretty small hassle—a requirement to choose among the health plans that I mentioned before—had a big impact on who got enrolled.
Second, what we saw was that it wasn’t an evenly distributed reduction in enrollment. It was bigger for some groups than others. The biggest reductions in enrollment were among young and healthy people, which makes sense because a hassle is something you have to work to do. And if you’re young and healthy, you might feel like the extra work isn’t worth it given that you don’t use much health care regularly.
We also saw that there was a bigger reduction in enrollment in health insurance among people who were living in low-income neighborhoods, living in high-minority neighborhoods.